do-git-worktree
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes standard shell utilities like
mkdirandbasename, along withgit worktreecommands. These actions are performed directly on the local filesystem in a designated path to facilitate branch management as described in the skill's purpose. - [INDIRECT_PROMPT_INJECTION]: The skill defines an attack surface by incorporating user-defined branch names into shell commands. This risk is addressed by the use of double-quoting for variables (e.g.,
"$branch","$dir") in the command templates, which prevents basic command injection via shell metacharacters. - Ingestion points: Untrusted user input for branch names and repository-derived metadata.
- Boundary markers: Not explicitly defined in the prompt templates.
- Capability inventory: Directory creation (
mkdir), repository path resolution (git rev-parse), and worktree manipulation (git worktree add/remove/prune). - Sanitization: The instructions follow shell scripting best practices by quoting variables to prevent word splitting or command execution from maliciously crafted branch names.
Audit Metadata